Best posts made by bbigford

@scottalanmiller said in Getting DHCP BAD_ADDRESS on Windows DHCP:

Have a DHCP server that is getting something trying to register to it with a BAD_ADDRESS. Obviously, we can't just look up the MAC address and track it down, because the MAC we get is bad.

We've got one new switch that went in about the time that the issue started, but we think that that is off of the network now, and the issue continues. There are loads of new web cams on the network, but they seem to all be working fine.

Any guesses on tracking down the issue?

Any time I've ran into this issue, it was because either someone statically assigned an address which resides in the DHCP scope (rather than doing a reservation, or using a different range). The other thing I've noticed is two devices, unaware of eachother, serving DHCP requests; such as a Windows Server with a DHCP role, and a Cisco router, both doing DHCP in the same range (human error obviously).

@bbigford said in Getting DHCP BAD_ADDRESS on Windows DHCP:

@scottalanmiller said in Getting DHCP BAD_ADDRESS on Windows DHCP:

Have a DHCP server that is getting something trying to register to it with a BAD_ADDRESS. Obviously, we can't just look up the MAC address and track it down, because the MAC we get is bad.

We've got one new switch that went in about the time that the issue started, but we think that that is off of the network now, and the issue continues. There are loads of new web cams on the network, but they seem to all be working fine.

Any guesses on tracking down the issue?

Any time I've ran into this issue, it was because either someone statically assigned an address which resides in the DHCP scope (rather than doing a reservation, or using a different range). The other thing I've noticed is two devices, unaware of eachother, serving DHCP requests; such as a Windows Server with a DHCP role, and a Cisco router, both doing DHCP in the same range (human error obviously).

I also went through a few forums and the two conflicting devices is what I've found to be the most common. One case I found was nefarious, so they had to configure dhcp-snooping on their networking device(s).

@scottalanmiller said in Namecheap site - cert expired:

@dbeato said in Namecheap site - cert expired:

I assume they renewed because now it just shows right....

I'm sure that they heard about it very quickly.

It expired on the 6th, they changed it on the 17th, but the cert didn't show as legit until the 18th. :face_with_tears_of_joy:

@scottalanmiller said in Namecheap site - cert expired:

@bbigford said in Namecheap site - cert expired:

@scottalanmiller said in Namecheap site - cert expired:

@dbeato said in Namecheap site - cert expired:

I assume they renewed because now it just shows right....

I'm sure that they heard about it very quickly.

It expired on the 6th, they changed it on the 17th, but the cert didn't show as legit until the 18th. :face_with_tears_of_joy:

Wow, just.... wow.

Oh and you couldn't login when it was expired. So their console was just offline for weeks. Any customers currently on them I've been offloading.

@gjacobse said in Win10 vs Fedora 28: Boot speed:

@bbigford said in Win10 vs Fedora 28: Boot speed:

@gjacobse said in Win10 vs Fedora 28: Boot speed:

@scottalanmiller said in Win10 vs Fedora 28: Boot speed:

From what I've seen, Windows 10 is the fastest. Microsoft really has high speed booting down to a science. I think they beat everyone at that. Fedora seems quite a bit faster than Ubuntu, but both are very slow compared to Windows 10 on the same hardware.

Sadly, just a place where Windows kicks butt over any Linux that I've seen. Not the biggest deal, but certainly a place where Fedora or Ubuntu could improve.

That is a bit frustrating considering the 'push' for getting away from MS and all of it's applications. I do know that I can improve performance some by putting in a SSD drive over the 7200RPM (SR) that's in it now. and I will at some point. Working on the priority list first.

5 minutes??? God, is your drive failing I wonder...

An SSD will get it down substantially. I want to say my Samsung 850 Pro boots Fedora nearly as fast as Windows 10; they are seconds apart, and that is around 5-15 seconds.

860 Pros came out not too long ago. Here's a 512GB if you can fit it in the budget; 256GB is cheaper of course.

https://www.amazon.com/Samsung-512GB-V-NAND-Solid-MZ-76P512BW/dp/B07836C6YV/ref=sr_1_1?ie=UTF8&qid=1541290811&sr=8-1&keywords=860+pro

Thanks for the link. That is definitely in a good price range.

Worth noting that I've used many other high end SSDs (Plextor, Intel, etc). Samsung manufacturers their own NAND, controller; everything, top to bottom. Theirs absolutely fly compared to any others I've used. For consumer use of course, not infrastructure.

This is a new one to me. Got a new CloudKey with a Unifi switch and nanoHD WAPs. CloudKey would allow me to login to it, but I would get a 'connection refused' when trying to access the controller. Powered it on today and was greeted with a boot loop. I put it in recovery mode and manually updated the firmware; that resolved the boot loop.

But doing an IP scan verified that the fallback IP is being used. When logging into the fall back IP, I just get 'connection refused' from any updated browser and using both HTTP/S.

Any ideas?

@scottalanmiller said in ubnt CloudKey - refused connection:

Might be hardware failure.

I was thinking the same thing. Putting the SD card into another CK did nothing, so I was thinking maybe a bad SD card. I did a factory reboot and left it sitting for about a half hour while I was doing some stuff and filling out an RMA. Suddenly it started performing exactly as it's supposed to. Going to keep my eye on it but it's been solid the rest of the day. Weirdness though; this is only the second CK I've had issues with (another one spontaneously turned off after running for about an hour and refused to come back on).

Re: ubnt CloudKey - refused connection

Forking the dicussion...

@JaredBusch said in MSP Ubiquiti Unifi management:

UniFi is entirely designed to be multi-site and multi-tenant.

Spin up a a UniFi controller on Vultr or wherever.
Go into CloudKey #1 and migrate
Go to Cloud instance and import the backup from the migration.
Complete the migration.
Your devices will just show up because that is how the Migration works.
Add specific users to this site only in the controller. Even Clients!

Now repeat the process from the next client.
Add access to new and existing users as appropriate.

That sounds pretty sweet. I'll take a look into doing this starting next week.

So with some SaaS, the client has their own tenant and then I just setup delegated access for us to be able to manage their tenant. I'm assuming it's similar, rather than a "site" being treated like a physical location, or extension to a core instance like a hub-spoke of an office headquarters and branch offices?

I am getting stumped on this one. I was cross-checking some other environments to make sure co-workers environments are up-to-snuff (pretty standard... fresh set of eyes).

I noticed there are 3 environments with on-prem Exchange that don't automatically redirect from HTTP to HTTPS. They require SSL, and I looked at issue tracking and we do get a lot of tickets saying "email doesn't work" with a simple response of "you have to put HTTPS". So, I went down the road of fixing that.

All 3 will be on Office 365 Q2 next year, but I'd like to fix it in the meantime. The 3 environments are Exchange 2010, 2013, and 2016 (all on the latest possible CUs).

Worth noting that I've tried this using various systems on different networks and also using various browsers, thinking maybe it was something on the endpoint side (cached, browser-related, etc).

The most complete guide I've found (albeit, it's 2010) and through a ton of threads finding that it still applies to 2013/2016 is this... https://www.vspbreda.nl/nl/exchange/1409/

I've also gone to C:\Program Files\Microsoft\Exchange Server\V14\ClientAccess\Owa\web.config and tried adjusting the following line: <httpRedirect enabled="false" destination="https://mail.domain.com/owa"; childOnly="true" />

Anyone got any ideas?

Edit: I'm not receiving an error page, where I'd need to do an absolute link in the Error Pages section of IIS; nonetheless, I had tried that anyway.

Thin client, absolutely. They are slow as fuck in most environments as they are just terrible hardware with an onboard OS that still needs to be patched. They are neither a stand-alone computer with full functionality, or a zero client with speed and security; they are the worst of both worlds.

Zero clients though, completely different story as that's a software-delivery discussion.

Microsoft only recently changed the "recommended" install to 64-bit.

Hardware acceleration has given me heartache in the past.

Disable add-ins, then enable one by one to see which one causes a hang (if any are causing that).

Run a memtest, I've (rarely) had this cause the issue but sometimes a bad stick from the factory has caused Excel-heavy users to experience performance issues.

On older PCs, it's sometimes due to a failing drive; but it's unlikely the SSD from the factory is the issue. Just worth noting.

Beyond that, it's often content based... pulling tons of links from the network, updating various content on the network that's using cells in the sheets, etc.